Introduction
Oil wastes generated during processing, transportation, and refining of
petroleum are serious environmental threats, especially in
petroleum-producing regions (Liu et al., 2009; Wang et al., 2012). These
wastes contain oily components, water, and mineral fractions, which can
include naturally occurring radioactive elements such as thorium, potassium,
radium, and others (Abo-Elmagd et al., 2010; Bakr, 2010). Yearly, about
60 million tons of oily wastes are generated (Hu et al., 2013). About
30–40 % of the oil wastes are radioactive; thus, this type of waste is
very common (Al-Masri et al., 2004; Hamlat et al., 2001; Selivanovskaya et al., 2013).
The waste materials are hazardous to plants, animals, and microorganisms due
to the presence of toxic and mutagenic compounds and their interactions
(Marin et al., 2005; Verma et al., 2006). In Russia, these wastes are
usually disposed of on the soil surface along the roads, around the new
industrial buildings and buildings under construction, etc.
(Galitskaya et al., 2014; Selivanovskaya et al., 2012). When disposed
of on the surface soil and exposed to precipitation, components of the oil
wastes can leach into the soil, altering the chemical, physical, and
biological properties (Mikkonen et al., 2012). As oil wastes are mixtures
of inorganic and organic compounds which can degrade to metabolites of
unknown persistence and toxicity, chemical quantification is insufficient to
estimate the environmental risk (Morelli et al., 2005; Mikkonen et al.,
2012).
Microorganisms are an essential part of terrestrial ecosystems, playing
important roles in soil biogeochemical cycles (Marcin et al., 2013; Li et
al., 2013). Soil microbial properties appear to be good indicators of soil
pollution, as they are very responsive and provide information about the
changes occurring in soil (Marin et al., 2005; Tejada et al., 2008). Soil
microbial biomass and basic respiration are the two parameters that are
traditionally used to estimate soil quality, particularly for soils polluted
by hydrocarbons (Labud et al., 2007; Lee et al., 2008; Lamy et al., 2013).
Another microbial parameter which can sensitively reflect the quality of
soils is microbial enzymes, as they participate in the biological cycling of
elements and the transformation of organic and mineral compounds
(Marin et al., 2005).
Changes in abiotic and biotic ecological factors significantly affect the
structure of bacterial and fungal soil communities; therefore, these changes
can be used as a tool for soil impact assessment (Huang et al.,
2013). To investigate the microbial community, shifts in soils,
culture-independent molecular techniques such as clone libraries, gradient
gel electrophoresis, single strand polymorphisms, terminal restriction
fragment length polymorphism, deep sequencing, and quantitative real-time
polymerase chain reaction are used (Adetutu et al., 2013; Bacosa et al.,
2012; Liu et al., 2013).
The effects of crude oil and oil waste on soil and its microbial community
have been studied (Lee et al., 2008; Labud, 2007; Marin et al., 2005;
Admon et al., 2001), while fewer publications are devoted to the hazards of
naturally occurring radioactive elements (Abo-Elmagd et al., 2010; Hrichi
et al., 2013) or their effects on bacteria (Zakeri et al., 2012).
Hydrocarbons can cause direct toxic effects on microbial cells due to their
ability to change fluidity and permeability of cell membranes and to alter
cell homeostasis, to inhibit enzymes, to disrupt the electron transport
chain and oxidative phosphorylation, and to cause lipid proliferation
(Ruffing and Trahan, 2014). Besides, hydrocarbon may cause indirect effects
on soil bacteria by changing aeration and water regimes. Radionuclides
may cause chromosomal aberrations, single strand breaks and base pair
substitution in the DNA of microorganisms (Min et al., 2003). The combined
effects of wastes, consisting of both heavy fraction hydrocarbons and
radionuclides, on soil still need to be investigated.
We hypothesized that the oil wastes disposed of on soil surfaces affect the
microbial communities due to both hydrocarbons and radioactive elements
contained in them. To assess these effects, column experiments were
performed. Raw (containing oily compounds and radionuclides) and treated
waste (containing mainly radionuclides) samples from a petroleum production
yard were investigated. The effects of total petroleum hydrocarbons (TPH) and radioactive elements on three
soil layers in columns (0–20, 20–40, and 40–60 cm) were investigated to
characterize: (a) the rate of migration of these contaminants, (b) the effects
on the microbiological characteristics of the soil layers (metabolic
coefficient and enzyme activities), and (c) shifts in the structure of
bacterial communities by means of polymerase chain reaction – single strand
conformation polymorphism (PCR-SSCP).
Materials and methods
Experimental design
In the experiment we used six soil columns of
60 cm × 10 cm × 10 cm (height × length × width) with undestroyed native soil
(Luvisol, Corg= 1.2 %, Ntot= 0.11 %, Kext= 91 g kg-1, Pext= 125 g kg-1) collected from the Matyushenski
forest nursery, Tatarstan, Russia (latitude: 55∘48′07′′ N,
longitude: 49∘16′13′′ E). Two columns were not artificially
contaminated by waste samples and served as a control (C-columns). On the
top of the other four columns we disposed of two waste samples (each waste
sample in two replicates), and thus the soil of these columns was considered
to be contaminated.
Sixteen waste samples were collected from tanks, pipes, and production
equipment in different seasons between 2010 and 2012 at the Tikchonovskii petroleum
production yard (Tatarstan, Russia; latitude: 54∘50′26′′ N,
longitude: 52∘27′08′′ E). Two of these waste samples were used
for analysing the toxicity and in the soil column experiment: a raw waste
sample (H) collected from tanks during cleaning and maintenance, and a
treated waste sample (R) obtained from equipment for oil waste treatment.
The TPH of the H-sample contained 36 % aromatics, 27 % alphaltenes,
16 % aliphatics, and 21 % resins. Thermal steam treatment is used in the
production yard to reduce the oil content. The quantity of waste samples H
and R loaded onto soil columns was calculated to equalize the activity
concentrations of 226Ra (about 1 kBq kg-1; H-columns and
R-columns, correspondingly). Over 30 days, the waste samples were situated
on the top of the soil columns and the rainfall was simulated based upon
the average atmospheric precipitation for the European part of Russia (650 mm a year).
After a month at 25 ∘C, soil from each column was divided into three
parts (upper layer: 0–20 cm (u), middle layer: 20–40 cm (m), and lower
layer: 40–60 cm (l) to give soil samples Hu, Hm, Hl, Ru, Rm, Rl, and Cu,
Cm, Cl) and analysed.
Chemical parameters
The total petroleum hydrocarbon (TPH) content in waste and soil samples was determined
by IR-spectrometry with an AN-2 analyser (LLC NEFTEHIMAVTOMATIKA-SPb, Saint
Petersburg, Russia). Fractionation of TPH into aromatics, aliphatics,
asphaltenes, and resins was done by silica gel column chromatography
followed by gravimetric analysis (Walker et al., 1975). TPH
extracts were dissolved in n-pentane and separated into soluble and
insoluble fractions (asphaltene). The soluble fraction was loaded on the top
of a silica gel G (60–120 mesh) column (2 cm × 30 cm) and eluted
with solvents of different polarities. The alkane fraction was eluted with
100 mL of hexane and then the aromatic fraction was eluted with 100 mL of
toluene. The resin fraction was eluted with 100 mL of methanol and
chloroform (Mishra, 2001).
Samples were dried for 24 h at 110 ∘C, homogenized, and sieved
through a 0.8 mm mesh. The sieved samples were weighed, packed in a
Marinelli-type beaker (1000 mL), sealed, and stored for 4 weeks to reach
equilibrium between 226Ra and its decay product. Gamma-ray
spectrometric measurements for natural radioactivity (226Ra,
232Th, and 40K) were performed with a Progress gamma spectrometer
(SPC Doza, Zelenograd Moscow, Russia) using a scintillation block for
detection based on a crystal of sodium iodide (Fotiou et al.,1998) at a
resolution of 30 keV at the 662 keV Cs-137 gamma line.
The total organic carbon content in waste samples was estimated according to
ISO 10694:1995, the total nitrogen content according to ISO
11261:1995, pH according to ISO 10390:2005, and
electroconductivity according to ISO 11265:1994.
Microbiological analysis
Soil metabolic quotient (qCO2) was calculated as the ratio of basal
microbial respiration to soil microbial biomass (Anderson and
Domsch, 1990). Basal respiration rates were determined according to
Schinner et al. (1995), and microbial biomass according to ISO
14240-2 (1997).
The dehydrogenase (DHA) activity of microorganisms was determined according
to the method described in Garcia et al., (1997). Soil (1 g) adjusted
to 60 % water-holding capacity was treated with 0.2 mL of 4 %
2-p-iodophenyl-3-p-nutrophenyl-5-phenyltetrazolium chloride and incubated at
22 ∘C in darkness (autoclaved soil samples were used as
controls). After 20 h, the iodonitrotetrazolium formazan (INTF) was
extracted with 10 mL of ethylene chloride / acetone (2:3), measured
spectrophotometrically at 490 nm, and the results were expressed as mg INTF
g-1 dry soil h-1.
Cellulase activity (CA) was estimated by hydrolysis of
carboxymethylcellulose according to the method described in Pancholy
and Rice, (1973) with modifications: soil (3 g) adjusted to 60 % water
holding capacity, 7.5 mL of 1.15 M phosphate buffer, 5 mL of 1 %
carboxymethylcellulose, and 0.5 mL of toluene were incubated at 28 ∘C for 24 h. The samples were filtered and 2 mL of
dinitrosalicylic acid reagent (10 g of 3.5-dinitrosalicylic acid, 16 g of
NaOH, and 300 g of K-Na-tartrate tetrahydrate in 1 L of distilled water) was
added to 4 mL of filtrate. The samples were then incubated at 95∘ N for
10 min in a water bath, cooled, and measured at 540 nm. Results were
expressed as milligrams of reducing sugars in 1 g of dry soil.
Single strand conformation polymorphism
Soil samples were sieved (4 mm mesh) and homogenized, DNA was extracted
using the FastDNA® SPIN Kit for Soil (Bio101, Qbiogene,
Heidelberg, Germany) according to the instructions provided, and the DNA
concentration was measured at 260 nm (Thermo Scientific GENESYS 20™,
Thermo Fisher Scientific Inc., Waltham, USA). DNA extracts were stored at
-20 ∘C for further analysis. Extraction was performed twice for
contaminated and control samples.
SSCP fingerprinting of the bacterial communities was performed as described
by Kampmann et al. (2012). Briefly, a PCR was performed (MyCycler, Bio-Rad, Munich, Germany) in a total volume
of 50 µL using chemicals and enzymes purchased from Fermentas (St.
Leon-Rot, Germany). The reaction mixture contained 0.6 of 0.02 µL-1 Dream Taq DNA Polymerase, 5 µL of 1 × Taq
Buffer, 4 µL of 2 mM MgCl2, 5 µL of 0.2 mM of each dNTP,
1 µL of 0.2 µmol-1 of each primer, 1 µL of 0.16 mg mL-1 BSA, and 2 µL of DNA. Bacterial communities were analysed
using the universal bacterial 16S rRNA gene primer pair Com1/Com2 (CAG CAG
CCG CGG TAA TAC / CCG TCA ATT CCT TTG AGT TT) (Schwieger and Tebbe,
1998) purchased from Eurofins MWG Operon (Ebersberg, Germany). The PCR
parameters were 95 ∘C for 3 min, followed by 16 cycles at 94 ∘C for 30 s, 64–57 ∘C for 30 s, and 72 ∘C
for 30 s, followed by 9 cycles at 94 ∘C for 30 s, 57 ∘C for 30 s, and 72 ∘C for 30 s, with a final elongation step of
30 min at 72 ∘C. PCR products were purified using the QiaQuick
PCR Purification Kit (Qiagen, Hilden, Germany). Before electrophoresis,
ssDNA fragments were generated by lambda exonuclease digestion according to
Schwieger and Tebbe (1998). The ssDNA was separated using the
INGENYphorU electrophoresis system (Ingeny International BV, Goes,
Netherlands) at 450 V and 19.5 ∘C for 17 h in a non-denaturing
polyacrylamide gel consisting of 0.6 × MDE solution (Biozym
Scientific GmbH, Hessisch Oldendorf, Germany) and 1 × TBE buffer
(0.89 M Tris, 0.89 M boric acid, and 20 mM EDTA pH 8.0). The gel was
silver-stained using the Page Silver Staining Kit (Fermentas, St. Leon-Rot,
Germany) according to the instructions provided and scanned to obtain
digitized gel images.
Identification of excised bands
Dominant bands were excised from SSCP gels as described by Schwieger and
Tebbe (1998). The gel-extracted DNA was reamplified and cloned as
described by Kampmann et al. (2012) using the
pGEM-T® Vector System (Promega, Mannheim, Germany). The four
clones of each band to be sequenced (LGC Genomics GmbH, Berlin, Germany)
using the M13 (Promega, Mannheim, Germany) forward primer were sent to LGC
Genomics GmbH (Berlin, Germany) in a 96-well microtiter plate filled with LB
(Lysogeny Broth)-Agar with 50 µg mL-1 of ampicillin.
Quality checks and cutting of sequences were performed using the software
package MEGA version 5.0 (Tamura et al., 2011). Sequences were analysed
for chimeras with the Pintail programme (Version 1, Cardiff School of
Biosciences, Cardiff, United Kingdom; Ashelford et al., 2005), and
putative chimeras were removed from the data set. Alignments were done with
the SILVA web aligner (SINA v1.2.11, Microbial Genomics and Bioinformatics
Research Group, Bremen, Germany; Pruesse et al., 2007), and similarity
values were calculated using the PHYLIP neighbour-joining algorithm
(Felsenstein, 1989) implemented in the ARB software package
(Ludwig et al., 2004). For sequence comparison, the SILVA SSU 106 Ref
database was used. Sequences were deposited in the NCBI GenBank database
with the accession numbers KF926419-KF926433.
Statistical analysis
Sampling and chemical analyses were carried out in triplicate and biological
analyses in quintuplicate, and all results were expressed on an air-dried
soil basis. Random variability of data was analysed to determine the mean
values and standard errors (S.E.). Statistical analyses were performed using
Origin 8.0 (OriginLab, Northampton, USA) and R Statistical Software (R
3.0.0, R Foundation for Statistical Computing Version, Vienna, Austria; R Development Core Team, 2012) packages.
SSCP gels were scanned at 400 dpi and the number of SCCP bands and their
areas and integrated intensities were estimated with Quantity One 1-D
Analysis Software (Biorad, Hercules, CA, USA). Each band was used as the
measured unit of biodiversity. Microbial community diversity was expressed
using several indices: Shannon-Weaver (H-index) and Simpson (D-index)
indices were calculated according to Shannon and Weaver, (1963) and
Simpson, (1949), respectively; the species diversity (S-index)
corresponded to the number of species in the line; the simple index
(I-index) was calculated as the number of bands in the SSCP line divided by
the number of bands in the line with the highest number of bands estimated
according to Silvestri et al., (2007); and the equitability of
the bands was calculated by Shannon's evenness (E-index; Zornoza et al.,
2009).
Two-way ANOVA with interaction was used to analyse the impact of factors
(e.g. depth of soil layers or type of contaminant) on the presence of bands
and microbial community diversity indices, and results yielding a p value
less than 0.01 were considered highly significant (Chambers and
Hastie, 1992). In all ANOVA, the number of degrees of freedom was two for
the type of contaminant, two for the depth of soil layers, four for the
interaction of these two factors, and nine for the residuals. The
F-statistic was in the range [6, 17.6] (p value range [0.02, 0.001]) for the
type of contaminant, [0.7596, 11.5] (p value range [0.5, 0.003]) for the
depth of soil layers, and [1.9, 5.1] (p value range [0.2, 0.02]) for the
interaction of these two factors. To visualize the differences in microbial
communities, metric multidimensional scaling (MDS) plots were created, where
matrices of band abundance were assembled, and similarity matrices were
calculated according to the Bray–Curtis coefficient (Faith et al.,
1987).
Cluster analysis was performed using hierarchic clusterization based on a
matrix of microbial communities dissimilarity. The Ward minimum variance
method from the Vegan package of the R software (R Foundation for
Statistical Computing Version 3.0.0, Vienna, Austria; R Development Core
Team, 2012), which aims to find compact, spherical clusters, was implemented
for clusterization (Ward, 1963).
Results and discussion
Chemical characterization of the wide range of waste samples
Oil wastes can contain radioactive elements and hydrocarbons in various
concentrations (Lazar et al., 1999). In our work, we estimated
TPH content and activity concentrations of 16 oily wastes sampled at a
petroleum production yard. As shown in Table 1, the TPH content ranged from
1.6 to 880.3 g kg-1, the activity concentration of 226Ra ranged
from 0.03 to 7.92, that of 232Th ranged from 0.02 to 5.09, and that of
40K ranged from 0.03 to 2.28 kBq kg-1. The values obtained are
comparable to or slightly exceed values reported by other authors (Liu et
al., 2009; Ros et al., 2010; Gazineu and Hazin, 2008; El Afifi and Awwad,
2005; Ayotamuno et al., 2007).
Chemical properties of the oil wastes from the Tikchonovskii
petroleum production yard.
Waste sample
TPH,
Activity concentration, kBq kg-1
number
g kg-1
226Ra
232Th
40K
1
35.0 ± 7.0
7.93 ± 1.62
2.40 ± 1.88
not detected
2
59.0 ± 11.8
0.62 ± 0.14
0.35 ± 0.07
not detected
3
90.4 ± 18.1
1.70 ± 0.37
0.30 ± 0.06
0.26 ± 0.04
4
880.3 ± 176.8
0.07 ± 0.02
0.02 ± 0.01
0.03 ± 0.01
5
95.4 ± 19.1
1.81 ± 0.39
0.28 ± 0.06
0.26 ± 0.04
6
720.1 ± 144.3
2.74 ± 0.60
0.92 ± 0.18
0.27 ± 0.04
7
123.3 ± 24.6
0.03 ± 0.01
0.03 ± 0.01
0.06 ± 0.01
8
57.4 ± 11.5
0.28 ± 0.06
0.15 ± 0.03
0.05 ± 0.01
9
59.2 ± 11.8
0.25 ± 0.05
0.11 ± 0.02
0.06 ± 0.01
10
30.5 ± 6.1
0.43 ± 0.10
0.20 ± 0.04
0.14 ± 0.02
11
46.5 ± 9.3
1.48 ± 0.33
0.12 ± 0.03
0.06 ± 0.01
12
153.1 ± 30.6
0.47 ± 0.10
0.25 ± 0.05
0.06 ± 0.01
13* (further H)
575.2 ± 121.0
4.40 ± 0.97
2.85 ± 0.57
1.28 ± 0.19
14* (further R)
1.6 ± 0.2
7.92 ± 0.93
3.99 ± 0.44
1.79 ± 0.21
15**
640.1 ± 128.3
3.86 ± 0.20
3.39 ± 0.08
1.27 ± 0.04
16**
4.6 ± 0.9
7.86 ± 1.73
5.09 ± 1.02
2.28 ± 0.34
* Pair of wastes in which waste sample no. 13 is raw waste and waste
sample no. 14 is the waste obtained by steam treatment of waste sample no. 13.
** Pair of wastes in which waste sample no. 15 is raw waste and waste
sample no. 16 is the waste obtained by steam treatment of waste sample no.
15.
The waste pairs 13/14 and 15/16 marked in Table 1 represent the two pairs of
untreated and treated waste samples. The treatment of these wastes, which is
a thermal steam treatment with chemical agents, is part of the industrial
process. The goal of the treatment is to reduce the hazardous properties of
the wastes.
For further investigation, we have chosen wastes 13 (further H) and 14
(further R) for the following reasons: (i) from the waste samples studied,
the initial waste sample H possesses quite a high concentration of TPH and,
at the same time, high activity concentrations of radionuclides; (ii) from
the waste samples studied, the treated waste sample R possesses the highest
activity concentration of 226Ra and the second highest activity
concentrations of 232Th and 40K; (iii) the composition of the
mineral part of the R-sample is the same as that of the H-sample, so the
effects of removing hydrocarbons from the waste can be studied.
Chemical characterization of the waste samples <inline-formula><mml:math display="inline"><mml:mi>H</mml:mi></mml:math></inline-formula> and <inline-formula><mml:math display="inline"><mml:mi>R</mml:mi></mml:math></inline-formula>
As shown in Table 1, the TPH content in sample H was estimated to be
575.2 ± 121.0 g kg-1, which is typical for this waste
(Ayotamuno et al., 2007; Al-Futaisi et al., 2007; Tahhan and Abu-Ateih,
2009; Selivanovskaya et al., 2013). The other physico-chemical
characteristics of the wastes were determined as follows: the distribution
of fractions in the H-sample was 26 ± 2 % asphaltenes, 23 ± 1 % resins, 19 ± 1 % aliphatics, and 32 ± 2 % aromatics.
EC in this sample was estimated to be 4.78 ± 0.56, and the pH was 7.2 ± 1 0.1. The C : N ratio was equal to 187 (TOC 747 ± 32,
Ntot 4 ± 0.2 g kg-1).
The treatment of the H-sample decreased the TPH content to 1.6 ± 0.2 g kg-1 (Table 1) and increased the activity concentrations of 226Ra,
232Th, and 40K 1.8-, 1.4-, and 1.8-fold, respectively; these
values are comparable with those reported by El Afifi and Awwad, (2005);
Bakr, (2010); Al-Saleh and Al-Harshan, (2008); Abo-Elmagd et al., (2010).
226Ra was the predominant isotope at 4.40 ± 0.31 kBq kg-1.
226Ra belongs to the uranium and thorium decay series, and the
awareness of radium isotopes is caused by the fact that it decays into radon
(222Rn), which is a Class A carcinogen (Zakeri et al., 2012). A
comparison of the results with the recommended IAEA levels for natural
radionuclides (IBSS, 2001) indicated that the waste samples could cause
environmental changes, as the values were 2.1- to 2.8-fold higher than
recommended for 232Th and 1.3- to 2.3-fold lower than recommended for
226Ra. Zakeri et al., (2012) reported that stress of 6
kBq or more from 226Ra influences growth characteristics, and stress of
1 kBq or more up-regulates proteins in a Serratia marcescens strain isolated from a hot spring.
In the R sample we observed the following fractions in TPH: 36 ± 3 %
asphaltenes, 33 ± 2 % resin, 12 ± 1 % aliphatics, and 19 ± 1 2 % aromatics. The electroconductivity of this sample was equal to
5.13 ± 0.4, the pH was 7.1 ± 0.1, and the C : N ratio was 35 (TOC:
2.10 ± 0.2; Ntot: 0.06 ± 0.01 g kg-1).
Chemical characteristics of the soil samples
The oil waste sample H and its treated variant R were added to the soil
columns once, and then over 30 days an amount of water equal to the yearly
local precipitation was added. In control columns, the TPH content and
activity concentration of radioactive elements were typical for the natural
soils (Starkov and Migunov, 2003; Vera Tomé et al., 2002; Shawky et
al., 2001; Gumerova et al., 2013), and did not change significantly between
the upper (0–20 cm), middle (20–40 cm), and lower (40–60 cm) layers. The
TPH content ranged from 0.2 to 0.4 g kg-1, the activity concentration
of 226Ra ranged from 0.01 to 0.02, that of 232Th ranged from 0.021
to 0.023, and that of 40K ranged from 0.29 to 0.34 kBq kg-1; these
values are within the worldwide averages (UNSCEAR, 2000).
Higher TPH and radionuclide content values were seen for H- and R-columns in
comparison to corresponding controls (Fig. 1), which indicated leaching of
toxic compounds from the waste samples into soil layers. H waste
samples increased the TPH content in H-soil columns, which was not observed
for R-columns. In the Hu samples, TPH content was estimated to be 3.5-fold
higher than in the corresponding control (Nu), while lesser amounts of
hydrocarbons had migrated into the middle and lower soil layers (2.8- and
2.2-fold greater than control). The trend for TPH distribution in soil
layers indicated that TPH contamination of deeper soil layers was to be
expected.
FExcess TPH content and activity concentrations of radionuclides in
soil sampled from H-columns (contaminated by the raw waste containing oily
compounds and radionuclides) and R-columns (contaminated by treated waste
containing mainly radionuclides) in different layers [upper (0–20 cm; u),
middle (20–40 cm; m), and lower (40–60 cm; l)] above the corresponding
values of the control columns.
![]()
Analyses of radionuclide activity concentrations indicated that
concentrations of 40K in the soil samples of H- and R-columns did not
differ from control values. In soil, this natural radioactive element
predominated, and the concentration was not high in the waste samples. The
average migration of other elements did not exceed 0.8 %. Presumably, the
leakage of 40K from waste samples was comparable with that of other
radionuclides, and therefore its migration did not change the natural level
of this radionuclide in soil samples.
Activity concentrations of 226Ra and 232Th were increased in
H-soil samples and 1.2- to 6.2-fold in R-soil samples over the control. The
activity concentrations of these two elements were much higher in Ru- and
Rm-samples compared to Hu- and Hm-samples, despite the fact that the amount
of waste added to the columns was equalized with respect to the activity
concentrations of radionuclides. Likely, radionuclides in raw waste samples
were part of organic complexes, which hindered their leakage into soil
layers with precipitation, while radionuclides migrated freely with water in
the R mineral sample.
Overall, it was shown that only low amounts (up to 0.8 %) of TPH and
radionuclides leaked into soil. But these relatively low concentrations did
alter the microbial community of soil as shown below.
Microbial community in soil samples
Soil metabolic quotient, cellulase, and dehydrogenase activities
Soil metabolic quotients (qCO2), which were expected to be higher in the
soil samples with higher microbial stress (Marin et al., 2005), are
presented in Fig. 2a. The lowest qCO2 was observed for the upper and
middle layers of the control columns, while the highest values were found
for the upper and middle layers of the H-columns and in the lower layers of
all three columns, where the microbial community was affected by oxygen and
organic matter limitations. The first is probably due to the effects of
hydrocarbons leached from the oil wastes on microorganisms.
Microbial characteristics of the soil sampled from columns C
(uncontaminated soil), H (contaminated by the raw waste containing oily
compounds and radionuclides), and R (contaminated by treated waste
containing mainly radionuclides) in different layers [upper (0–20 cm; u),
middle (20–40 cm; m), and lower (40–60 cm; l)]. A – metabolic quotient
(qCO2), B – cellulase activity (CA), C – dehydrogenase activity (DHA).
![]()
Cellulases are important enzymes in the carbon cycle, and CA may be used to
indicate soil impacts (Sinegani and Sinegani, 2012). As shown in
Fig. 2b, CA in all soil columns decreased from the upper to the lower
layers. No significant differences were found between R-samples and
corresponding control samples, but in Hu and Hm samples, CA was
1.4-fold lower than that of the controls, which indicated that hydrocarbons
can decrease the cellulase activity of the soil.
DHA is often used as a parameter for the estimation of soil quality, in
particular for the hydrocarbon degradation rate (Margesin et al., 2000;
Marin et al., 2005). In this study, no significant correlation between DHA
and the toxic element content or soil depth was found for R- and H-samples
(from all three layers; Fig. 2c). This disagreed with the results reported
in Lee et al., (2008) and Tejada et al., (2008), where a
significant negative correlation between TPH content in soil and DHA was
seen. However, these authors worked with soils containing 4.5–100 g kg-1 of hydrocarbons, whereas in this study, TPH levels did not exceed
1.3 g kg-1.
Microbial parameter values for H- and R-columns were 62, 70, 95, and 80, 95,
110 % of corresponding control samples for the upper, middle, and lower
layers, respectively. These data indicated that the highest stress existed
in Hu and Hm samples, which were influenced by the raw waste. Radionuclides
appeared to play a less important role for microbial functional properties.
Microbial community structure
Shifts in microbial community structure are sensitive indicators for
assessing the changes in soils under the influence of pollution as well as
other biotic and abiotic factors. Recently, culture-independent methods were
used to estimate the number of strains belonging to different ecological or
systematic groups (Adetutu et al., 2013). In this study, PCR-SSCP was
used to describe the changes in microbial community structure (Schwieger and
Tebbe, 1998).
Bacterial species identified after sequencing of bands obtained from SSCP
gels
Total bacterial DNA was extracted, amplified by PCR using common bacterial
primers for 16S rDNA, and separated by polyacrylamide gel electrophoresis
(SSCP profiles are shown in Fig. 3). SSCP patterns demonstrated variations
between different soil layers and types of contaminants (oily components +
radionuclides or only radionuclides), where 21 to 34 discrete bands of
various intensities were observed for each SSCP line, and the types of
bands were identified using relative electrophoretic distances. In total,
488 bands were detected and 25 were observed in at least two independent
SSCP profiles.
SSCP profiles of the bacterial communities of soil sampled from columns C (uncontaminated soil), H (contaminated by the raw waste
containing oily compounds and radionuclides), and R (contaminated by treated
waste containing mainly radionuclides) in different layers [upper (0–20 cm; u), middle (20–40 cm; m), and lower (40–60 cm; l)].
![]()
Selected bands 1–4 (Fig. 3) were excised from the gel, cleaned, cloned, and
sequenced. Bands 1 and 2 were considered stable, as they were present in all
samples (except Ru) in relatively large amounts. Band 3 dominated in control
samples, and its relative abundance expressed in terms of area and band
intensity was 1.5- to 3.8-fold higher in control samples than in H- and
R-samples. Band 4 was present in all samples, but was predominant in
H-samples.
Four randomly-picked clones of each band (after blue-white selection) were
sequenced, and the next relatives were identified by a similarity matrix
using the neighbour-joining algorithm implemented in the ARB software and the
SILVA database SSU 106 Ref. Clones of band 1 (KF926419-KF926422) were
phylogenetically similar to Burkholderia strains found in unpolluted and polluted sites
(AF247491, DQ465451, FJ210816; Weisskopf et al., 2011; Friedrich et al.,
2000), while clones of band 2 (KF926423-926425) were similar to strains of
Burkholderia and Bradyrhizobium jicamae (JX010967, JN662515). Bacteria from the genus Burkholderia are typical soil
inhabitants, and certain Burkholderia strains are resistant to hydrocarbons and are used
in the bioremediation of oil-polluted sites (Bacosa et al., 2012;
Weisskopf et al., 2011; Hamamura et al., 2008; Adetutu et al., 2013). Band 3
(KF926426-KF926429), which is sensitive to oily and radioactive components
in the waste samples (not seen in contaminated H- and R-samples), was
genetically similar to Hydrogenobacter hydrogenophilus (Z30242) uncultured Acidobacteria isolated from unpolluted
grassland and forest soils (HQ598830, HQ599021; Naether et al., 2012)
and an uncultured Chlorobiales bacterium found in a uranium mining waste pile (AJ295649,
AJ536877; Selenska-Pobell, 2002). Band 4 (KF926430-KF926433), which
dominated in H-columns, was related to an uncultured bacterium from mineral
soils of the Atacama desert (JX098489, JX098426; Lynch
et al., 2012) and actinomycetes from the genus Catenulispora (CP001700, AJ865857; Busti et al., 2006) as well as strains isolated from
gasoline-polluted sites (or able to degrade hydrocarbons; JQ919514; Hilyard et al., 2008), including a Parvibaculum strain that catabolizes
linear alkylbenzene sulfonate (AY387398; Schleheck et al.,
2004).
Continued.
![]()
SSCP analysis of PCR products and statistical analysis
The microbial diversity of each sample was calculated using five indices,
and the results are presented in Table 2.
The biodiversity indices of the soil sampled from the upper (u),
middle (m), and lower (l) layers of the control (C), raw waste (H), and
treated waste (R) contaminated columns.
Samples
S
I
H
D
E
Cu
25
0.74
3.01
0.94
0.93
Cu
29
0.85
3.05
0.94
0.90
Cm
33
0.97
3.31
0.96
0.95
Cm
34
1.00
3.38
0.96
0.96
Cl
31
0.91
3.19
0.95
0.93
Cl
26
0.76
2.91
0.93
0.89
Hu
29
0.85
2.80
0.89
0.83
Hu
24
0.71
2.79
0.91
0.88
Hm
23
0.68
2.84
0.92
0.90
Hm
29
0.85
3.11
0.94
0.92
Hl
25
0.74
2.84
0.92
0.88
Hl
25
0.74
2.83
0.92
0.88
Ru
28
0.82
3.22
0.96
0.97
Ru
27
0.79
3.17
0.95
0.96
Rm
29
0.85
3.09
0.94
0.92
Rm
29
0.85
3.24
0.96
0.96
Rl
21
0.62
2.72
0.91
0.89
Rl
21
0.62
2.73
0.91
0.89
H-index – Shannon-Weaver index, D-index – Simpson index, S-index – number
of species (bands) in SSCP profile, I-index – simple index, E-index –
Shannon's evenness (index of equitability of the bands).
The S-index represented the number of SSCP bands in a line (in the sample).
The number of bands ranged between 25 and 34 in C-columns, between 23 and 29
in H-columns, and between 21 and 29 in R-columns. No significant differences
were seen between samples from H, R, and control columns with respect to
depth. Only in the Rl-samples did the number of the SSCP patterns decrease
significantly in comparison to samples from the upper and middle layers. The
average number of bands tended to be higher in the control samples (29.7)
compared to the contaminated samples (25.8 each). According to the data
presented in the literature, the influence of combined hydrocarbon and
mineral contamination of soil can lead to both increases and decreases in
its microbial diversity. Thus, the increase of microbial diversity is
explained by the fact that TPH can be used by microorganisms as carbon
sources. Therefore, a relatively low TPH input could lead to the development
of a new hydrocarbon-degrading species without suppression of indigenous microbes
(Gao et al., 2015; Nie et al., 2009). Negative effects on soil biodiversity are explained
by significant inhibition of indigenous microflora in the oil-contaminated
sites because of the toxic influence of hydrocarbons or their metabolites,
oxygen deficit, and other factors (Hui et al., 2007; Morelli et al.,
2005; Marcin et al., 2013).
The I-index reflected the diversity of bands in the sample with respect to
the sample with the highest biodiversity; the highest I-indices were
observed in Cm samples and the lowest in Rl samples. The community diversity
Shannon-Weaver index (H-index), which is expected to be higher in samples
with the highest number of bands but with similar frequencies, fluctuated
from 2.72 to 3.38. This reflected the variety of band profiles among
samples, which indicated changes in the microbial community due to waste
compounds or depth. The evenness (E-index) was higher in the samples with
higher H-indices (R= 0.86). The Simpson D-index was smaller when one band
predominated, and the lowest D-indices were observed for Rl and Hu samples.
The sample compositions differed significantly between H-samples and other
samples, and the bands labelled 4 (Fig. 4) were dominant while band 3
disappeared. These results were in agreement with those of Morelli et al. (2005), who observed that organisms in polluted ecosystems which
are capable of degrading contaminants or resisting toxicity are dominant,
while other species do not survive.
Cluster analysis of the SSCP bands observed on SSCP profiles of
soil sampled from the columns C (uncontaminated soil), H (contaminated by
the raw waste containing oily compounds and radionuclides), and R
(contaminated by treated waste containing mainly radionuclides) in different
layers [upper (0–20 cm; u), middle (20–40 cm; m), and lower (40–60 cm; l)].
![]()
The ANOVA of the linear model of influence of factors (type of waste, depth,
their interdependence, and residuals) on biodiversity was performed. The
presence of oil waste was significant only for the D-index (p < 0.01), while other indices of biodiversity did not depend on the factors
investigated.
The correlation between factors describing soil samples (type of waste,
depth, their interdependence, and residuals) and microbial community
structure was examined. ANOVA of the presence or absence of 25 bands (which
were observed in at least two samples) was carried out, and it was found
that depth was a significant factor for five bands, the presence of
contaminant for six, and the combined influence of these two factors for
three (p < 0.01). An ANOVA for the MDS values was performed as
suggested by Lin et al., (2012) to reduce the dimensions of the values
analysed. The type of waste, as well as the interaction between waste and
depth, was significant for the structure of the microbial community. Depth
did not play an important role in the bacterial community structure (p < 0.01) and the control columns did not differ between soil layers,
as opposed to H- and R-columns.
Samples were grouped using MDS and clustering analysis methods. Cluster
analysis, which orders samples according to their similarity indices, is
commonly used to show the differences or classification between groups of
clusters (Kadali et al., 2012). To determine the number of
clusters on the dendrogram (Fig. 4), the method of natural break was
implemented. The samples were divided into two groups: the first group
included all control samples, while the second group contained R-samples
from the upper and middle layers as well as all H- and R-samples from lower
layers (the samples of the second group contained fewer microbial strains).
The first group was subdivided into three parts according to the type of
waste or depth: Rm-samples, C-samples of the upper and middle layers
(further subdivided into Rl-samples, Hl-samples, and H-samples of the upper
and middle layers), and Ru/Cl samples.
MDS is the most common ordination method used for ecological community data
(Wilson et al., 2013; Terahara et al., 2004). Figure 5 shows the MDS plot
(r2= 0.56 for distance correspondence), where the closer to one
another the points representing microbial communities were situated on the
plot, the more similar these microbial communities were. Samples were
positioned according to the type of contaminant (H, R, and uncontaminated
control, C), which could be explained by the selective influence of toxic
compounds from H and R on the strains present in soil. This finding is
consistent with that of Hamamura et al. (2008), who suggested
that the population shifts corresponding to the prominent bands in soils are
due to the content of hydrocarbons. It is important to note that communities
from the R-columns were separated from the communities from C-columns,
despite the fact that the activity concentration of 226Ra was below the
recommended level (IBSS, 2001) and not in line with the estimates for
functional characteristics of the microbial community. This confirmed that
PCR-based estimates of environmental influence can be more sensitive than
traditional methods (Lin et al., 2012; Bialek et al., 2011).
Metric multidimensional scaling analysis based on distance matrix
of SSCP profiles of soil sampled from the columns C (uncontaminated soil),
H (contaminated by the raw waste containing oily compounds and radionuclides),
and R (contaminated by treated waste containing mainly radionuclides) on
different layers [upper (0–20 cm; u), middle (20–40 cm; m), and lower
(40–60 cm; l)].
![]()
Conclusions
Oil wastes generated during processing, transportation, and refining of
petroleum, which are frequently disposed on the soil surface, are serious
environmental threats, especially in petroleum-producing regions. In this
study, we have investigated the combined effects of hydrocarbons and
radionuclides contained in oil waste on the soil microbial community. Such
effects have not been studied before, although a large amount of oil waste
is radioactive. We analysed the wastes from tanks, pipes, and production
equipment sampled in different seasons between 2010 and 2012 and established a wide
range of TPH content from 1.6 ± 0.2 to 880.3 ± 176.8 g kg-1, activity concentration of 226Ra from 0.03 ± 0.01 to
7.92 ± 0.93, activity concentration of 232Th from 0.02 ± 0.01 to 5.09 ± 1.02, and activity concentration of 40K from
0.03 ± 0.01 to 2.28 ± 0.34 kBq kg-1. To distinguish between
the effects of hydrocarbons and radionuclides, we chose the raw waste H with
a typical content of TPH and radionuclides and its treated variant with
reduced hydrocarbon content but containing radionuclides (waste R).
The sample H contained 4.40 ± 0.31 kBq kg-1 of 226Ra, 2.85 ± 0.21 kBq kg-1 of 232Th,
and 575.2 ± 121.0 g kg-1 of TPH and the sample R contained 7.9 ± 1.8 kBq kg-1 of
226Ra, 3.9 ± 0.9 kBq kg-1 of 232Th, and 1.6 ± 0.4 g kg-1 of TPH. The last two compounds exceeded the levels
reported to be non-toxic in the environment, indicating that the traditional
practice where oil waste was spread on the soil surface could have negative
effects on the soil.
Disposal of H waste samples on the soil surface increased the TPH content in
H-soil columns, which was not observed for R-columns. In the soil sampled
from the upper layer of the H-column, the TPH content was estimated to be
3.5-fold higher than in the corresponding control sample, while lesser
amounts of hydrocarbons had migrated into the middle and lower soil layers
(2.8 and 2.2 times higher than control). Despite the fact that the amount of
waste samples disposed of on the tops of soil columns was equalized
according to the amount of 226Ra, a greater amount of this radionuclide
was observed in the soil of R-columns: it was 4.3, 1.4, and 1.2 times higher
than that in H-columns in the upper, middle, and lower layers, respectively.
It is likely that radionuclides in raw waste samples were part of organic
complexes which hindered their leakage into soil layers with precipitation,
while radionuclides migrated freely with water in the R mineral sample.
By analysing the functional characteristics of soil microorganisms, oil
compounds (but not radionuclides) were found to influence soil microflora.
The qCO2 and cellulase activity in soil samples from H-columns were
reduced 1.3 to 2.2 times more than in R-columns, where microbial activity
values were close to the control values. In contrast, PCR-SSCP demonstrated
that both oil compounds and radioactive elements could cause shifts in the microbial
community structure.
We conclude that oil waste containing radioactive elements caused negative
changes of soil microbial community by its disposal, while petroleum
hydrocarbons played the more pronounced negative role. The effects of
radionuclides contained in oily waste on soil can be evaluated using
culture-independent analyses of microbial communities.